scrapecreators

When To Use

Use this skill when the user needs one comprehensive public social-media dossier that deliberately spans Twitter/X, LinkedIn, Instagram, and TikTok. It retrieves public profile and content evidence for one coherent target and adds a relevant TikTok hashtag plus a region-specific trending baseline.

This is a fixed, paid bundle—not an endpoint menu. Do not use it for one simple profile or post lookup; choose the smallest matching endpoint or a narrower skill instead. It does not retrieve protected/private content, reveal contact details, send messages, post, follow, like, or otherwise mutate an account.

Workflow

1. Install the vetted recipe:

   selat skill install scrapecreators

2. Collect all nine required inputs and verify that they describe one coherent research target:

   • twitterHandle and tweetId;
   • linkedinProfileUrl, linkedinPostUrl, and linkedinCompanyUrl;
   • instagramHandle;
   • tiktokHandle, tiktokHashtag, and region.

   Handles may differ across platforms. Remove leading @ and # characters. Confirm that the tweet and LinkedIn post are public and relevant to the same subject; do not substitute unrelated examples or placeholders.

3. Probe every payment challenge for free:

   selat skill verify ~/.config/selat/skills/scrapecreators \
     --twitterHandle "satyanadella" \
     --tweetId "1632748758613241857" \
     --linkedinProfileUrl "https://www.linkedin.com/in/satyanadella/" \
     --linkedinPostUrl "https://www.linkedin.com/posts/satyanadella_its-been-a-busy-few-weeks-between-today-activity-7323480567562276865-F2mk" \
     --linkedinCompanyUrl "https://www.linkedin.com/company/microsoft/" \
     --instagramHandle "microsoft" \
     --tiktokHandle "microsoft" \
     --tiktokHashtag "microsoft" \
     --region "US" \
     --live-probe

4. Show the 11 current live quotes, their expected cumulative total, and the sum of all per-step caps. Wait for explicit approval of a cumulative session budget no higher than the cap sum. A free verification is not evidence that a particular public identity will return useful provider data.

5. After approval and a spendable Gateway balance, arm a session budget equal to the approved amount, run the fixed bundle once, and disarm the budget after success or failure. The session budget is a selat-pay spending tripwire across the session. It does not select or skip steps:

   selat budget start --amount <approved-cumulative-cap>
   selat skill run scrapecreators \
     --twitterHandle "<twitter-handle>" \
     --tweetId "<numeric-tweet-id>" \
     --linkedinProfileUrl "<linkedin-person-url>" \
     --linkedinPostUrl "<linkedin-post-or-article-url>" \
     --linkedinCompanyUrl "<linkedin-company-url>" \
     --instagramHandle "<instagram-handle>" \
     --tiktokHandle "<tiktok-handle>" \
     --tiktokHashtag "<hashtag-without-#>" \
     --region "<two-letter-region>"
   selat budget stop

The CLI executes every manifest step independently and continues after an individual failure. Inspect per-step results and payment history; a partial failure does not mean earlier calls were uncharged. Never retry until history is checked, prices are re-probed, and the retry is separately approved.

Fixed Steps

1. Twitter/X public account profile.
2. Twitter/X recent public posts for the same handle.
3. Details for one supplied public tweet ID.
4. LinkedIn public person profile.
5. One supplied public LinkedIn post or article.
6. LinkedIn public company page.
7. Instagram public profile, including the provider's recent-profile context.
8. One page of public Instagram posts in trimmed form.
9. TikTok public profile metadata.
10. One page of public TikTok posts for a relevant hashtag and region.
11. A trimmed TikTok trending feed for the same region.

The calls do not pass data between steps. The agent must correlate identifiers, normalize returned records, and synthesize the final report after the run.

Inputs And Outputs

Return:

1. target identity and retrieval time;
2. a platform-by-platform evidence table with source URL, profile/content facts, observed engagement fields, and limitations;
3. cross-platform consistencies and conflicts;
4. TikTok hashtag and regional-trend context clearly separated from facts about the target account; and
5. per-step status plus final settled dollar cost.

Do not imply that the TikTok region filter limits results to creators located in that region: it selects the provider's proxy/feed context. Do not calculate engagement rates when the required denominator is absent, infer sensitive traits, or treat public availability as consent for outreach.

Rails And Costs

• The three SELAT-native Twitter calls currently resolve as routed-x402 via Circle Gateway at about $0.001 each.
• The eight synchronous Scrape Creators calls currently resolve as routed-mpp over MPP on Tempo at about $0.021 each.
• Every step requires a reachable SELAT_ROUTER_URL.

maxAmount is a per-call ceiling, not a price or cumulative run cap. The three Twitter caps are $0.002 each and the eight Scrape Creators caps are $0.03 each, for a cumulative cap sum of $0.246. The top-level $0.03 is only a fallback for a future step without its own override.

The free live probe on 2026-10-04 quoted an expected total of $0.171 (3 × $0.001 + 8 × $0.021). Always re-probe before payment because prices, rails, and endpoint health can change.

Gotchas

• Fixed pipeline. selat skill run executes and pays for all 11 calls on every run. There is no platform selector, step filter, or conditional branch, so a run cannot be scoped to a subset of platforms. For a one-platform need, hand-build the single selat-pay call instead (see references/endpoints.md).
• No duplicate calls. Each step hits a distinct endpoint. The two Instagram steps are profile and posts; the three LinkedIn steps take three different URL types.
• No inter-step dataflow. Returned user IDs, cursors, and URLs are not automatically inserted into later calls.
• Only the first page is included. Additional Instagram/TikTok pages are separate calls. Re-probe and obtain approval before pagination.
• Identifiers are platform-specific. Do not assume one handle is shared across Twitter, Instagram, and TikTok.
• LinkedIn URL types matter. Person, post/article, and company endpoints require their matching URL type.
• Public, read-only scope. Protected/private records and account mutations are out of scope.
• A paid application error may still charge. A step can settle and still return a provider error or empty payload. Check selat history before any retry, and never re-run the whole bundle to repair one failed step.

Validation

• Static: selat skill validate ./skills/scrapecreators
• Live gate, free: run the full selat skill verify command shown above with coherent public inputs and --live-probe.
• Paid verification: only after fresh quotes, explicit approval, and an armed cumulative session budget, add --pay; every call may settle independently.
• Provider request schemas and single-step free probes are documented in references/endpoints.md.

References

• manifest.json — machine-readable fixed payment recipe.
• references/endpoints.md — request schemas, current modes/prices, and QC corrections.
• ../../references/agent-skill-authoring-sop.md — authoring standard.
• selat-pay — https://github.com/SELAT-AI/selat-pay

Provider and product names are used only for endpoint identification. This skill is not affiliated with X, LinkedIn, Instagram, TikTok, or Scrape Creators.